All work
№ 07Shipped

NFL Draft Prediction

Ensemble prediction model built for the GCI Tokyo 2026 Data Science Programme.

What I built

  • Raised prediction AUC from a 0.808 baseline to 0.837 through a rank-average ensemble of LightGBM, XGBoost, CatBoost and Random Forest, with Optuna hyperparameter tuning, cross-validation and feature engineering.
  • Completed the GCI Tokyo 2026 Data Science Programme's AUC-scored competitive prediction track.

Interested?

Let's build on this

If NFL Draft Prediction is close to something you're working on, I'd be glad to talk — collaboration, contract work, or research.

Collaborate on this project

colleenmarasha01@gmail.com